Apparatus and method for scanning an object

CROSS-REFERENCE TO RELATED APPLICATIONS

0001The present application is a divisional of U.S. patent application Ser. No. 11/298,171 having a filing date of Dec. 10, 2005, now U.S. Pat. No. 7,706,028, which is a continuation-in-part of U.S. patent application Ser. No. 10/915,900 having a filing date of Aug. 11, 2004, now U.S. Pat. No. 7,557,966, the entire disclosures of which are incorporated herein by reference.

FIELD OF THE INVENTION

0002This invention relates generally to method for scanning an object. More particularly, this invention relates to a method for scanning the three-dimensional shape of a foot.

BACKGROUND OF THE INVENTION

0003The footwear industry has become a large, specialized, and competitive environment. The industry has long-established methods of operation, one of which is to determine shoe size based upon the length and width of the foot. Shoe lasts, which are solid forms upon which shoes are constructed for each shoe size, are used in the manufacture of shoes. Therefore, properly sized lasts are essential for a successful shoe line. Properly fitting shoes are always important to consumers, but fit can be particularly important for golfers who wear relatively stiff golf shoes over varied terrain for relatively long periods of time and in varied weather conditions. Golfers often experience blisters caused by poorly fitting golf shoes. Poorly fitting shoes can also affect the golfers' hitting performance by not providing a stable base for their feet during golf swings. Thus, it is important to obtain the best fitting shoes to minimize such problems.

0004Various mechanical techniques have been proposed in the past for obtaining foot measurements. For example, most shoe stores commonly use a foot measuring scale known as the Brannock device, produced by the Brannock Device Company of Syracuse, N.Y. This device consists of a metal base plate with several sliding scales. It measures the length and width of the foot to determine an appropriate shoe size. One problem associated with the Brannock device is that its foot measurements are only two dimensional in nature, measuring an absolute length from heel to toe and the width. This method fails to take into consideration factors such as type of heel, for example, bony versus padded; shape of the toes, for example, square versus tapered; insole arch; and other characteristics. This device also fails to measure characteristics associated with medical problems such as bunions, which require larger shoe sizes to accommodate abnormalities of the foot.

0005Some systems use cameras to determine the characteristics of the foot. U.S. Pat. No. 5,911,126 to Massen discloses a method and arrangement for digitizing three-dimensional sensing of the shape of bodies or body parts. The data can be used for automated fabrication of a shoe last, bedding or foot molding, for example. This system uses an elastic envelope worn over the foot/body part during imaging by several cameras surrounding the foot/body part. The envelope carries a high-contrast pattern that allows for digitizing of the image. This method requires that a high contrast pattern is applied onto the bodies or body parts to provide the contrast for digital manipulation.

0006Other systems have proposed the use of laser beam measurement to determine the characteristics of the foot, as disclosed in U.S. Pat. No. 4,745,290 to Frankel et al.; U.S. Pat. No. 5,164,793 to Wolfersberger et al., and U.S. Pat. No. 5,237,520 to White. In the Frankel and Wolfersberger references, the heel area is not measured. In the White patent, the dimensions of the upper foot are not obtained. Since all areas of the foot are important for a proper fit, omitting these areas is undesirable. Additionally lasers, which are structured light sources, do not illuminate the entire foot at once. The foot is typically painted line-by-line, sequentially until the entire foot is painted. This method of illumination takes time and requires a control mechanism for the laser. The foot must also remain stationary during this time period. Furthermore, laser systems are expensive.

0007U.S. Pat. No. 6,633,326 to Fukumoto describes an image system, whereby an image pick-up head having a CCD camera mounted thereon captures image(s) of a foot. The image pick-up head is mounted on an elliptical rail surrounding the foot. The reference also discloses a particular arrangement of signal cable and auxiliary cable to prevent entanglement. The foot needs to remain stationary for a relatively long time for the camera to image all sides of the foot. This reference does not disclose how the image(s) can be processed to create a three-dimensional model of the foot.

0008Hence, there remains a need in the art for an apparatus and method for accurately measuring feet for production of lasts and selection of proper fitting shoes, among other uses.

0009Further, in obtaining an image from a scanning system, it is often desirable to position the object within a very specific field of view for optimal image capture. Hence, there remains a need in the art for an automatic object placement verification system for use with an apparatus and method for accurately measuring feet for production of lasts and selection of proper fitting shoes among other uses.

SUMMARY OF THE INVENTION

0010The present invention is directed to a method for establishing a position of an object to be scanned, comprising the steps of:

0011(i) placing the object on a scanning platform, wherein the scanning platform includes a predetermined scanning area having at least one sensor;

0012(ii) obtaining a signal from the sensor; and

0013(iii) analyzing the signal to determine if the object is within the predetermined scanning area.

0014The sensor may be positioned either outside of or within the predetermined scanning area. The sensor may comprise a pressure-resistive sensor. If the object is not positioned within the predetermined scanning area, a signal may be sent to a user interface to reposition the object. The interface may receive directional guidance for repositioning the object. In one version, the method involves determining if the object is positioned on a sensor or at least a portion of the object is placed on a plurality of sensors.

BRIEF DESCRIPTION OF THE DRAWINGS

In the accompanying drawings, which form a part of the specification and are to be read in conjunction therewith, like reference numerals are used to indicate like parts in the various views:

<figref idref="DRAWINGS">FIG. 1</figref> is a perspective view of a scanner in accordance with the present invention;

<figref idref="DRAWINGS">FIG. 2</figref> is a side view of the scanner is <figref idref="DRAWINGS">FIG. 1</figref>;

<figref idref="DRAWINGS">FIG. 3</figref> is a perspective view of the cameras and processing equipment contained within the scanner of <figref idref="DRAWINGS">FIG. 1</figref>;

<figref idref="DRAWINGS">FIG. 4</figref> is a top view of the cameras and processing equipment of <figref idref="DRAWINGS">FIG. 3</figref>;

<figref idref="DRAWINGS">FIG. 5</figref> is a cross-sectional view along line <b>5</b>-<b>5</b> in <figref idref="DRAWINGS">FIG. 4</figref>;

<figref idref="DRAWINGS">FIG. 6</figref> is a cross-sectional view along line <b>6</b>-<b>6</b> in <figref idref="DRAWINGS">FIG. 4</figref>;

<figref idref="DRAWINGS">FIG. 7</figref> is a front view of the printed circuit board (with certain details omitted for clarity) located behind one of the camera pairs illustrating the triangulation technique and calibration technique;

<figref idref="DRAWINGS">FIG. 7A</figref> is a schematic view showing an angle defined by the camera pairs and the object to be scanned;

<figref idref="DRAWINGS">FIG. 7B</figref> is a schematic view of overlapping FOV between the camera pairs;

<figref idref="DRAWINGS">FIG. 8A</figref> is an exemplary portion of a triangulated image in gray scale;

<figref idref="DRAWINGS">FIG. 8B</figref> is an exemplary portion of <figref idref="DRAWINGS">FIG. 8A</figref> in binary black/white scale;

<figref idref="DRAWINGS">FIG. 9</figref> is an exemplary point cloud representative of a scanned foot;

<figref idref="DRAWINGS">FIG. 10</figref> is a top view of an alternate embodiment of the cameras and processing equipment of a scanner in accordance with the present invention showing the sensors and processing equipment of an automated foot placement system;

<figref idref="DRAWINGS">FIG. 11</figref> is a flow chart showing an exemplary method for scanning a foot; and

<figref idref="DRAWINGS">FIG. 12</figref> is a flow chart showing an exemplary method for determining and/or correcting foot placement.

DETAILED DESCRIPTION OF THE INVENTION

0031Referring to <figref idref="DRAWINGS">FIG. 1</figref>, scanner <b>10</b> comprises a display or monitor <b>12</b> supported by neck <b>14</b> on housing <b>16</b>. Display <b>12</b> preferably has touch screen capability to interact with users. Neck <b>14</b> can be telescopic so that monitor <b>12</b> can be raised or lowered. Housing <b>16</b> defines cavity <b>18</b>, which is sized and configured to receive an object to be scanned. Any three-dimensional object can be scanned, and cavity <b>18</b> can be enlarged to accommodate objects of any size. The present invention is described below using a foot; however, the present invention is not limited to any particular scanned object.

0032As shown in <figref idref="DRAWINGS">FIG. 2</figref>, housing <b>16</b> is preferably orientated at an incline relative to a surface supporting the scanner. The inventors of the present invention have determined that a foot inclined at a predetermined range of inclination angles exerts even pressure on the foot. In other words, at the preferred inclination the foot exerts even pressure on the heel and the balls. Advantageously, no part of the foot is unevenly deformed or squashed so as not to cause improper foot measurements. For persons weighing between 100 lbs and 250 lbs, the preferred inclination angle ranges from about 12° to about 20°, more preferably about 14° to about 18°, and most preferably about 16°.

0033Referring to <figref idref="DRAWINGS">FIGS. 3-6</figref>, inside housing <b>16</b> and supported on platform <b>20</b> are camera pairs <b>22</b> comprising cameras <b>24</b>, <b>26</b>, and central processing unit or equipment (CPU) <b>28</b>. Platform <b>20</b> may also support a hard drive, input/output (I/O) boards, USB boards, telephone modem, wireless communication devices, display units, and the like. Camera pairs are preferred, because the pairs provide depth perception to the image, similar to binoculars and human/animal eyes. The processes for manipulating the data from the images taken from the camera pairs are discussed below. Although the present invention is described herein with cameras <b>22</b> arranged in pairs, single cameras can also be used, instead of pairs with the present invention. For single cameras, depth perception can be obtained by using adjacent cameras as pairs. A single camera can be a part of two pairs with its two adjacent cameras. A single camera can also form a pair with a non-adjacent camera. Alternatively, a single camera may be rotatably mounted on a support such that the camera may be toggled between two positions so as to simulate having a pair of cameras. For example, a camera mount may be electronically controlled to take a picture in a first orientation, move the camera a specified distance, such as through a predetermined angle in any direction, to a second orientation, and then take a second picture in the second orientation. Preferably, the specified distance is small, such as less than about 5 degrees.

0034Unique to the present invention, unstructured light is used to illuminate the scanned object covered by a textured surface. Unstructured light includes, but is not limited to, sunlight, diodes, fluorescent light, and incandescent light, among others. Unstructured light broadly includes any light sources that don't have controlled frequencies. The entire scanned object can be illuminated at once, thereby eliminating the time consuming step of sequentially painting an object line-by-line with a structured laser beam. The optical pattern necessary for analyzing the scanned object is provided by a textured surface covering the scanned object. In accordance with one aspect of the present invention, the textured surface can be a textile sock covering a foot. Preferably, the sock fits snugly over the foot and without the sag rings around the ankles, so that a proper foot measurement can be achieved. Suitable socks can have any color or pattern and preferably have one color. If the socks have multiple colors, preferably the colors do not substantially contrast each other. The socks can be athletic socks or dress socks, and can be made from woven, non-woven, stitch-bonded non-woven or knit textiles.

0035At least three camera pairs can be deployed to image the foot, including toe camera pair <b>32</b> and two side camera pairs <b>34</b>. These three camera pairs can image the top and sides of the foot. Alternatively, the two side camera pairs <b>34</b> can be used along with a positioning device, such as a heel stop or toe stop (not shown). Any number of cameras can be deployed, and preferably seven camera pairs are used, as shown in <figref idref="DRAWINGS">FIGS. 3-6</figref>.

0036As best shown in <figref idref="DRAWINGS">FIG. 4</figref>, one of the camera pairs is preferably toe camera pairs <b>32</b>, and the remaining camera pairs are preferably evenly distributed as shown to capture all surface areas of the foot. Two of the remaining camera pairs are preferably side cameras <b>34</b>. Alternatively, camera pairs <b>22</b> are not evenly spaced apart, so long as their fields of vision (FOVs) overlap each other so that all surface areas of the foot are imaged. Camera pairs <b>22</b> can be evenly, angularly spaced apart, while the distances to the scanned object are uneven, or vice versa. In accordance with another aspect of the present invention, the FOVs do not necessarily overlap each other, but each FOV images a distinct part of the foot to measure the size of the foot and/or to form a last representative of the foot. In the art of shoemaking, twelve separate measurements of the foot are necessary to construct a last. Hence, each camera or camera pair can measure one or more measurements to make a last.

0037To obviate the need to focus the cameras prior to imaging, the cameras preferably have relatively small apertures. The aperture can be as small as F-22. Smaller aperture provides a relatively long distance within which no focusing is necessary. In one embodiment, an F-8 aperture is selected for the cameras. This aperture provides a fixed focus from about 6 inches to 14 inches away from the lenses. Preferably, the scanned object is positioned within this fixed focus distance. Camera pairs <b>22</b> may have different aperture sizes.

0038Suitable cameras include, but are not limited to, CMOS (complimentary metal-oxide sensor) digital or electro-optical cameras. Suitable CMOS cameras are commercially available from OmniVision Technologies, Inc. of Sunnyvale, Calif.; Micron Technology, Inc., of Boise, Id.; and Veo, Inc. of San Jose, Calif., among others. Each of the cameras has a lens that direct light to strike an array or panel of image sensor photo-sensitive cells or pixels. Each photo-sensitive cell or pixel is a photo-transistor which converts light into a voltage signal representing the light level. The voltage signal is stored as analog information and then digitized by an analog-digital-converter. Alternatively, CCD (charge coupled device) cameras can be used. CMOS and CCD cameras are widely available as consumer digital cameras. Analog cameras can also be used, so long as the image can be digitized thereafter, for example, by digital scanning.

0039Referring to <figref idref="DRAWINGS">FIG. 5</figref>, where a front view of camera pairs <b>22</b> is shown, cameras <b>24</b> and <b>26</b> of camera pairs <b>22</b> are supported on front face <b>36</b>. Front face <b>36</b> also protects the electronics on the printed circuit board (PC) <b>38</b>. As shown in <figref idref="DRAWINGS">FIG. 7</figref> with other details omitted for clarity, directly behind each camera <b>24</b>, <b>26</b>, is the photo-sensitive panel <b>24</b>′, <b>26</b>′, respectively. Preferably, photo-sensitive panels <b>24</b>′, <b>26</b>′ are positioned parallel or co-planar to each other. According to one aspect of the invention, each photo-sensitive panel comprises a two-dimensional array of 1280 pixels by 1084 pixels to give each panel about 1.38 megapixels. In one embodiment, the cameras are black and white cameras, and each pixel, when exposed, records a gray value having a numerical representation of 0-255 on the gray scale. Alternatively, color cameras can be used. During manufacturing, each panel <b>24</b>′, <b>26</b>′ is fixedly attached to PC <b>38</b> by molten solder. Due to its relatively high viscosity, the surface tension of the molten solder keeps the panels oriented substantially parallel to PC <b>38</b>. As the solder hardens, photo-sensitive panels <b>24</b>′, <b>26</b>′ are fixed to PC <b>38</b>.

0040To compensate for any misalignment from this parallel arrangement, each PC <b>38</b> is calibrated. Referring again to <figref idref="DRAWINGS">FIG. 7</figref>, cameras <b>24</b>, <b>26</b> (not shown) are mounted on top of photo-sensitive panel <b>24</b>′, <b>26</b>′ in front of calibration board <b>40</b>. Calibration board <b>40</b> comprises a plurality of markings <b>42</b>. Each marking <b>42</b> is preferably unique from each other and the locations of markings <b>42</b> on calibration board <b>40</b> are known. Calibration board <b>40</b> is located parallel to PC <b>38</b> at known distances, so that the location of each marking <b>42</b> in three-dimensional space relative to cameras <b>24</b>, <b>26</b> is known. Each marking <b>42</b> is imaged by cameras <b>24</b>, <b>26</b> via focal points <b>24</b><i>f</i>, <b>26</b><i>f</i>, respectively. Each marking is exposed at different locations on photo-sensitive panels <b>24</b>′, <b>26</b>′. By comparing the locations of each marking on photo-sensitive panels <b>24</b>′, <b>26</b>′ to each other and to the known location of corresponding marking <b>42</b> on calibration board <b>40</b> using the triangulation technique discussed below, any misalignment from the parallel orientation can be detected. Calibration board <b>40</b> is preferably positioned at three different locations from PC <b>38</b>, for example, 6 inches, 11 inches and 14 inches away from PC <b>38</b>, during calibration. Corrections to restore the parallel arrangement can be calculated and stored on a memory on each PC <b>38</b>.

0041One readily apparent advantage of the present invention is that by placing both photo-sensitive panels and both cameras on a single PC board it is easier to arrange the photo-sensitive panels to be parallel or co-planar to each other. As used herein, the term parallel includes both parallel and co-planar.

0042As best shown in <figref idref="DRAWINGS">FIG. 5</figref>, the cameras in pairs <b>22</b> are offset with respect to each other, that is, they do not align vertically or horizontally with each other. In one aspect of the invention, cameras <b>24</b> and <b>26</b> are offset about 0.5 inches in the vertical direction and about 1.0 inches in the horizontal direction relative to platform <b>20</b>. Any offsetting geometry can be utilized, although the offset is preferably between about 0.5 inches and 1.0 inches in any direction. The offsetting location enhances the depth perception of the camera pairs.

0043The method of locating a single point or pixel in three-dimensional space by both cameras is based on binocular triangulation methodology. Referring again to <figref idref="DRAWINGS">FIG. 7</figref>, an arbitrary point “a” in three dimensional space (x, y, z) or (R, Ø, θ) creates image point <b>24</b><i>a </i>on panel <b>24</b>′ and image point <b>26</b><i>a </i>on panel <b>26</b>′. In one example, point “a” can be any point on the textured surface of the sock covered foot. The relative locations of these image points are different on the photo-sensitive panels. In other words, image points <b>24</b><i>a </i>and <b>26</b><i>a </i>are located at different relative positions on photo-sensitive panels <b>24</b>′ and <b>26</b>′. To match image points <b>24</b><i>a </i>to image point <b>26</b><i>a </i>and to represent point “a” in three-dimensional space, in one embodiment a PC processor, which is preferably located on PC <b>38</b>, searches and matches neighboring points or pixels in the vertical and horizontal directions. When the neighboring horizontal and vertical pixels match each other, then a match is found for image point <b>24</b><i>a </i>and image point <b>26</b><i>a. </i>When the contour of the foot or the scanned foot is sharp, for example, around the ankle or greater than 45°, or when point “a” is located near the edge of the camera lens, image points <b>24</b><i>a</i>, <b>26</b><i>a </i>may not match each other. Unmatched points in the images are relatively few in numbers, and some images do not have unmatched points.

0044Binocular triangulation also provides depth perception, because an object is viewed by two lenses (or eyes) and the two images are joined to create a single image. The depth or distance between the scanned object (for example, the scanned foot or calibration board <b>40</b>) and cameras <b>24</b>, <b>26</b> can be determined from the distance between matched points <b>24</b><i>a </i>and <b>26</b><i>a</i>, as measured on the PC board and the angle between line <b>24</b>″ connecting point <b>24</b><i>a </i>to point “a” and line <b>26</b>″ connecting point <b>26</b><i>a </i>to point “a.” Hence, by employing camera pairs the location of any pixel representing a point on the scanned foot in three-dimensional space relative to the camera pairs is known. Also, as discussed above binocular triangulation can be used with single cameras instead of camera pairs, when adjacent (or non-adjacent) single cameras are paired up for triangulation.

0045As shown in <figref idref="DRAWINGS">FIG. 7A</figref>, cameras <b>24</b> and <b>26</b> are positioned relative to each other such an incident angle, α, is defined between image point <b>24</b><i>a </i>and image point <b>26</b><i>a </i>and point a on the object to be scanned. Preferably, incident angle a is less than about 90 degrees and preferably less than about 45 degrees so that the overlap <b>25</b>, shown in <figref idref="DRAWINGS">FIG. 7B</figref> of the FOV of camera <b>24</b> and the FOV of camera <b>26</b> is maximized. The overlap of these fields of view is preferably greater than about 60%, and, more preferably, at least about 80%, although any amount of overlap may be used in the present invention. Overlapping FOVs and incident angle a are exemplary factors that can be utilized to control the offset or positions of camera pairs.

0046To minimize calculation errors, after image point <b>24</b><i>a </i>is matched to image point <b>26</b><i>a</i>, the distance between point “a” in three-dimensional space as seen by camera <b>24</b> and the same point “a” as seen by camera <b>26</b> is calculated. Optimally, this distance should be as small as possible, and preferably it should approach zero. If this distance is greater than a predetermined threshold, then the PC processor or CPU <b>28</b> should return an error signal. Alternatively, a mid-point is chosen based on this distance and point “a” is set or reset to the mid-point. For the purposes of this application, “mid-point” is considered to be the shortest distance, or the common point, for a camera pair <b>22</b>. Any arbitrary point within this distance can be used, for example, one-third, one-fourth, one-fifth or any fraction of this distance, can be used.

0047As stated above, each PC <b>38</b> preferably has its own processor or controller to process the data generated by the cameras and a memory storage device (such as flash memory, EEPROM, EPROM, RAM, etc.) to store the raw data as well as the triangulated data of matched points and unmatched points, to store the calibration routine, and to transfer the triangulated data to CPU <b>28</b> for further analysis. PC <b>38</b> can communicate with CPU <b>28</b> by radio frequency, serial or parallel port connections, optical fibers, USB, USB-2, among others. USB and USB-2 are the preferred connection modes due to their high transmission speeds and low costs.

0048After calibration, each PC and camera pair are fixedly attached to platform <b>20</b>, such that the positions of the camera pairs are known and should not change significantly. Each camera pair can image other camera pairs in its FOV to calibrate and measure the relative distance between the cameras. For example, referring to <figref idref="DRAWINGS">FIG. 4</figref>, left side camera pair <b>34</b> can image right side camera pair <b>34</b> and lower right hand camera pair <b>22</b> can image upper left hand camera pair <b>22</b> and toe camera pair <b>32</b> to fix their positions. The relative positions of the camera pairs to each other and on platform <b>20</b> can be stored in CPU <b>28</b>.

0049Since each camera pair <b>22</b> has been calibrated so that the positions in three-dimensional space of the scanned image in its FOV relative to the camera pair are known, and since the locations of each camera pair relative to each other on platform <b>20</b> in scanner <b>10</b> are also known, the locations of each pixel in three dimensional space representing the scanned foot after the images from each camera pair have been stitched together are also known.

0050After CPU <b>28</b> receives all of the triangulated images captured by the camera pairs, CPU <b>28</b> stitches or otherwise combines these images to create a three-dimensional image of the sock-covered foot, e.g., point cloud. An exemplary image received by CPU <b>28</b> is shown in <figref idref="DRAWINGS">FIG. 8A</figref>. In one preferred embodiment, an area <b>43</b> equivalent to 32 pixels by 32 pixels is selected for matching an image to adjacent images. Selected area <b>43</b> represents a portion of the textured sock imaged by a camera pair. Selected areas <b>43</b> are located in the areas where the FOVs of adjacent camera pairs are overlapping each other. In any selected area <b>43</b>, each pixel has a gray scale value from 0 to 255 produced by the black and white digital cameras. Zero value can designate white and a 255 value can designate black and everything in between represents a different shade of gray. Hence, a selected area <b>43</b> in gray scale in one image can be searched and matched to a corresponding selected area <b>43</b> on an adjacent image. Once a match is made, two adjacent triangulated images can be stitched together with the matched selected areas <b>43</b> superimposing each other.

0051Since selected area <b>43</b> in gray scale requires a significant amount of memory storage and processing speed, efficiency and processing speed can be improved when selected area <b>43</b> is converted from gray scale to black/white or binary scale. A predetermined threshold gray value is selected. Any gray scale value above this threshold gray scale is converted to black and any gray scale value below this gray scale is converted to white, as shown in <figref idref="DRAWINGS">FIG. 8B</figref>. After the conversion, one or more islands <b>44</b> are created in selected area <b>43</b>. The predetermined threshold can be the average gray scale or can be any value less than 255. When the threshold is low, more islands are created and when the threshold is high, fewer islands are created. The threshold value can also be adjusted while scanner <b>10</b> is in operation to assist with the image stitching process. Islands <b>44</b> in selected area <b>43</b> can be used to match an image to adjacent images. To minimize calculation errors, during the island matching process the size and location of islands <b>44</b> are compared. If the size difference or the location difference exceeds predetermined thresholds, then CPU <b>28</b> would also return an error message.

0052During the stitching process, unmatched points from one triangulated image are searched for in adjacent triangulated images. If the unmatched image points in one image have corresponding matched image points in adjacent images, then the matched points are used. If too many unmatched points remain, CPU <b>28</b> may discard the images and instruct the camera pairs to take another set of images of the foot.

0053From the seven images captured by the seven camera pairs <b>22</b>, point cloud <b>48</b> representing the scanned foot can be generated after the seven images are stitched together, as shown in <figref idref="DRAWINGS">FIG. 9</figref>. As described above, since the relative and fixed positions of camera pairs <b>22</b> on platform <b>20</b> in scanner <b>10</b> are known, and since the images taken by the camera pairs also contain three-dimensional positions of each pixel relative to the cameras, point cloud <b>48</b> is a three-dimensional representation of the scanned foot.

0054Alternatively, instead of a point cloud representation a convolution filtering process can be applied to the stitched image to provide an embossed view of the foot. Convolution filtering is commonly used in commercially available Photoshop software and is typically used to provide a clearer outline of the scanned object.

0055As will be recognized by those in the art, camera pairs <b>22</b> may alternatively include three or more cameras to form camera clusters. Each camera cluster would be positioned in scanner <b>10</b> as camera pairs <b>22</b>, as described above. The binocular triangulation method described above could be achieved by using signals obtained from any two cameras within a camera cluster. For example, a camera cluster may include three cameras, A, B, and C. Cameras A and B could be used to obtain one triangulation point, Cameras B and C could be used to obtain a second triangulation point, and/or Cameras A and C could be used to obtain a third triangulation point. Obtaining several readings from each camera cluster could help to assure the accuracy of the measurements.

0056Referring to <figref idref="DRAWINGS">FIG. 10</figref>, an exemplary process of scanning both feet is illustrated. Either the right foot or the left foot can be scanned first. The first step is to adjust the light level in cavity <b>18</b> (see <figref idref="DRAWINGS">FIG. 1</figref>) where the foot is inserted. Each camera or camera pair has its own light sensor. Scanner <b>10</b> may have its own unstructured light sources, which can be located between camera pairs <b>22</b> or located proximate to each camera pair. The amount of illumination generated by these light sources can be controlled by CPU <b>28</b> or by the PC controllers for the camera pairs depending on the level of light detected by the light sensors. In one embodiment, the light sources are attached to the underside of the roof of housing <b>16</b>.

0057The next steps are to check all the cameras and to determine whether the foot is properly placed in cavity <b>18</b>. If all of the foot, including toes and heel, can be imaged, then the foot positioning is acceptable. The placement of the foot can be checked by imaging the foot with some or all of the cameras. Scanner <b>10</b> may also optionally include a placement system <b>50</b> for determining the correct placement of objects within scanner <b>10</b>.

0058In one embodiment, as shown in <figref idref="DRAWINGS">FIG. 3</figref>, placement system <b>50</b> may be a visual guide. For example, a pad <b>30</b> is defined proximate to the center of platform <b>20</b> to provide visual cues for the correct or preferred placement of the foot. Pad <b>30</b> is preferably printed or colored to show a user a correct placement position.

0059In another embodiment, as shown in <figref idref="DRAWINGS">FIG. 10</figref>, system <b>50</b> may also be an automatic placement system that includes a plurality of sensors <b>51</b>, <b>52</b>, <b>54</b>, <b>56</b>, <b>56</b>′, <b>58</b>, and <b>58</b>′ for detecting the placement of the object to be scanned, a communication interface board <b>60</b> for gathering the information from sensors <b>51</b>, <b>52</b>, <b>54</b>, <b>56</b>, <b>56</b>′, <b>58</b>, and <b>58</b>′ and communicating the information to CPU <b>28</b>, and associated software disposed on a hard surface such as platform <b>20</b>. In one embodiment, platform <b>20</b> is made of aluminum, but it may be made of any rigid material, such as plastic, steel, or the like.

0060Sensors <b>51</b>, <b>52</b>, <b>54</b>, <b>56</b>, <b>56</b>′, <b>58</b>, and <b>58</b>′ are preferably pressure resistive sensors or sensor arrays that have low static resistance and high resistance when a force is applied thereto. For example, sensors <b>51</b>, <b>52</b>, and <b>54</b> are strip pressure resistive strip sensors and sensors <b>56</b>, <b>56</b>′ and <b>58</b>, <b>58</b>′ are pairs of pad sensors; all sensors are preferably permanently or removably affixed to platform <b>20</b>, such as with an adhesive, welding, or printing. As will be apparent to those in the art, any pad sensor pairs, that is, sensors <b>56</b>, <b>56</b>′ and <b>58</b>, <b>58</b>′, may optionally be replaced with a single large sensor. Sensors <b>51</b>, <b>52</b>, <b>54</b>, <b>56</b>, <b>56</b>′, <b>58</b>, and <b>58</b>′ are preferably thin to prevent interference with the scanning process. In one embodiment, sensors <b>51</b>, <b>52</b>, <b>54</b>, <b>56</b>, <b>56</b>′, <b>58</b>, and <b>58</b>′ have a thickness less than about 2 mm.

0061Sensors <b>51</b>, <b>52</b>, <b>54</b>, <b>56</b>, <b>56</b>′, <b>58</b>, and <b>58</b>′ are preferably positioned on platform <b>20</b> such that, as shown in <figref idref="DRAWINGS">FIG. 10</figref>, when a foot is improperly placed on platform <b>20</b>, sensor <b>52</b> encounters, a heel of a foot, that is, sensor <b>52</b> is powered ON, sensor <b>51</b> or sensor <b>54</b> encounters the sides of a foot, and sensor pair <b>58</b>, <b>58</b>′ encounters the toes. Similarly, when properly placed, sensor pair <b>56</b>, <b>56</b>′ encounters a ball of a foot. In a special scenario, if a very large foot is properly placed on platform <b>20</b>, sensor <b>51</b>, sensor <b>54</b>, and sensor pair <b>56</b>, <b>56</b>′ all encounter parts of a foot. For this exemplary configuration, Table 1 illustrates various scenarios with combinations of sensors powered on or off and whether or not this results in a scan or a failure.

0062<tables id="TABLE-US-00001" num="00001"><table frame="none" colsep="0" rowsep="0"><tgroup align="left" colsep="0" rowsep="0" cols="1"><colspec colname="1" colwidth="217pt" align="center" /><thead><row><entry namest="1" nameend="1" rowsep="1">TABLE 1</entry></row></thead><tbody valign="top"><row><entry namest="1" nameend="1" align="center" rowsep="1" /></row><row><entry>Exemplary Sensor Combinations</entry></row></tbody></tgroup><tgroup align="left" colsep="0" rowsep="0" cols="6"><colspec colname="1" colwidth="35pt" align="left" /><colspec colname="2" colwidth="28pt" align="left" /><colspec colname="3" colwidth="28pt" align="left" /><colspec colname="4" colwidth="42pt" align="left" /><colspec colname="5" colwidth="42pt" align="left" /><colspec colname="6" colwidth="42pt" align="left" /><tbody valign="top"><row><entry>Sensor</entry><entry>Sensor</entry><entry>Sensor</entry><entry>Sensor</entry><entry>Sensor</entry><entry /></row><row><entry>51</entry><entry>52</entry><entry>54</entry><entry>Pair 56, 56′</entry><entry>Pair 58, 58′</entry><entry>Result</entry></row><row><entry namest="1" nameend="6" align="center" rowsep="1" /></row><row><entry>OFF</entry><entry>OFF</entry><entry>OFF</entry><entry>ON</entry><entry>OFF</entry><entry>SCAN</entry></row><row><entry>ON</entry><entry>OFF</entry><entry>ON</entry><entry>ON</entry><entry>OFF</entry><entry>SCAN</entry></row><row><entry /><entry /><entry /><entry /><entry /><entry>(assume very</entry></row><row><entry /><entry /><entry /><entry /><entry /><entry>large foot)</entry></row><row><entry>OFF</entry><entry>OFF</entry><entry>ON</entry><entry>ON</entry><entry>OFF</entry><entry>FAILURE</entry></row><row><entry>ON</entry><entry>OFF</entry><entry>OFF</entry><entry>ON</entry><entry>OFF</entry><entry>FAILURE</entry></row><row><entry>OFF</entry><entry>OFF</entry><entry>OFF</entry><entry>ON</entry><entry>ON</entry><entry>FAILURE</entry></row><row><entry>OFF</entry><entry>ON</entry><entry>OFF</entry><entry>ON</entry><entry>OFF</entry><entry>FAILURE</entry></row><row><entry>OFF</entry><entry>OFF</entry><entry>OFF</entry><entry>OFF</entry><entry>OFF</entry><entry>FAILURE</entry></row><row><entry namest="1" nameend="6" align="center" rowsep="1" /></row></tbody></tgroup></table></tables>

0063As will be apparent to those in the art, additional permutations exist for even this exemplary embodiment, such as where multiple sensors are powered ON in combinations not shown in Table 1. In such cases, the scan will not initiate. Additional configurations of sensors on platform <b>20</b> are also possible, where the combination of sensors powered ON and sensors powered OFF that result in a scan are determined based upon the actual configuration. For example, sensor pair <b>56</b>, <b>56</b>′ may be eliminated, so that a scan will only initiate when no sensors are powered ON. In another example, only three sensors may be used, a toe sensor, a ball sensor and a heel sensor, where a scan will only initiate when all sensors are powered ON. In another example, all sensors could be placed very close together so that a scan will only initiate when all sensors are powered ON. All possible configurations of sensors are contemplated by the present invention.

0064Sensors <b>51</b>, <b>52</b>, <b>54</b>, <b>56</b>, <b>56</b>′, <b>58</b>, and <b>58</b>′ are connected to communications interface board <b>60</b> via connections <b>64</b>. Connections <b>64</b> may be any type of connection known in the art, such as electrical wires or filaments, printed leads or a wireless link transmitting information from a transmitter on sensors <b>51</b>, <b>52</b>, <b>54</b>, <b>56</b>, <b>56</b>′, <b>58</b>, and <b>58</b>′. Similar to sensors <b>51</b>, <b>52</b>, <b>54</b>, <b>56</b>, <b>56</b>′, <b>58</b>, and <b>58</b>′, connections <b>64</b> are preferably affixed to platform <b>20</b> using any method known in the art. Preferably, connections <b>64</b> are adhered to the surface of platform <b>20</b>, e.g. by adhesives, tapes, etchings, etc.

0065Communications interface board <b>60</b> includes a microprocessor <b>62</b> and at least one digital comparator circuit (not shown) for converting the analog signals generated by sensors <b>51</b>, <b>52</b>, <b>54</b>, <b>56</b>, <b>56</b>′, <b>58</b>, and <b>58</b>′ into digital information usable by microprocessor <b>62</b>. Preferably, each sensor has a dedicated digital comparator circuit. Communications interface board <b>60</b> is linked to CPU <b>28</b> using any method known in the art, preferably an electrical wire or filament such as connector <b>66</b>.

0066Microprocessor <b>62</b> may be any microprocessor known in the art capable of collecting the digital inputs from the digital comparator circuit(s) and communicating with CPU <b>28</b>. One such suitable microprocessor <b>62</b> is an IC MCU flash chip available from Microchip Technology, Inc. of Chandler, Ariz. The communication with CPU <b>28</b> may be achieved by any method known in the art; however, for ease of manufacturing and interchangeability, the communication with CPU <b>28</b> is preferably achieved using a network according to the RS-485 standard protocol.

0067A thin film of material having a visual graphic printed or otherwise disposed thereupon preferably covers system <b>50</b>, but at least covers sensors <b>51</b>, <b>52</b>, <b>54</b>, <b>56</b>, <b>56</b>′, <b>58</b>, and <b>58</b>′. The film may be made of any material known in the art, but is preferably a material such as Mylar® polyester having a non-smearing graphic printed thereupon.

0068To operate system <b>50</b>, CPU <b>28</b> hosts a system manager software program. This program hosts the user interface program, preferably a graphical user interface (GUI), and provides all of the hardware, software, and GUI interaction. This program may be any type of program capable of performing these functions, and it may be written in any code known in the art. For example, the system manager communicates with communications interface board <b>60</b> to signal microprocessor <b>62</b> to interrogate sensors <b>51</b>, <b>52</b>, <b>54</b>, <b>56</b>, <b>56</b>′, <b>58</b>, and <b>58</b>′, as well to retrieve the information from sensors <b>51</b>, <b>52</b>, <b>54</b>, <b>56</b>, <b>56</b>′, <b>58</b>, and <b>58</b>′ to be interpreted by CPU <b>28</b>.

0069Referring to <figref idref="DRAWINGS">FIG. 5</figref>, one exemplary process of scanning both feet is illustrated. Either the right foot or the left foot can be scanned first. The first step is to adjust the light level in cavity <b>18</b> (see <figref idref="DRAWINGS">FIG. 1</figref>) where the foot is inserted. Each camera or camera pair has its own light sensor. Scanner <b>10</b> may have its own unstructured light sources, which can be located between camera pairs <b>22</b> or located proximate to each camera pair. The amount of illumination generated by these light sources can be controlled by CPU <b>28</b> or by the PC controllers for the camera pairs depending on the level of light detected by the light sensors. In one embodiment, the light sources are attached to the underside of the roof of housing <b>16</b>.

0070To initiate a scan, a user inserts an object such as a foot into scanner <b>10</b> and inputs through the GUI a signal to begin the scanning process. When the system manager program receives this input, the system manager queries microprocessor <b>62</b> using any type of communication known in the art, such as communications according to the RS-232 standard protocol. The next steps are to check all the cameras and to determine whether the foot is properly placed in cavity <b>18</b>. The placement of the foot can be checked, for example, by following the process shown graphically in <figref idref="DRAWINGS">FIG. 6</figref>. First, the system manager sends a signal to the microprocessor to query sensors <b>51</b>, <b>52</b>, <b>54</b>, <b>56</b>, <b>56</b>′, <b>58</b>, and <b>58</b>′. If the foot is appropriately positioned within the scanning area, then the system manager allows the scan to begin. If the sensors show a resistance, that is, if the foot is placed on one of sensors <b>51</b>, <b>52</b>, <b>54</b>, <b>56</b>, <b>56</b>′, <b>58</b>, and <b>58</b>′ such that the foot is outside of the appropriate field of view of the camera pairs, then the system manager sends a signal to the GUI to instruct the user to reposition the foot in a certain direction, such as to move the foot forward, backward, or to the left or right or using any other method as known in the art.

0071After the foot position is confirmed, the foot can be re-imaged by all the camera pairs, or the first set of images can be fully processed. If the foot is improperly positioned, then the user is instructed to move his/her foot before the foot can be imaged.

0072After the images are taken by camera pairs <b>22</b>, the images are analyzed to ensure their qualities, for example, whether the number of unmatched image points are acceptable or whether the distance between the same point “a” seen by camera <b>24</b> and camera <b>26</b> is within acceptable range. Then the images can be stitched to form a point cloud.

0073Next, the CPU determines whether the point cloud is acceptable, for example, whether selected areas <b>43</b> between adjacent images can be matched. Then, the point cloud is checked to determine whether it resembles a foot.

0074After these steps are satisfactorily executed, the other foot is scanned and the same steps are repeated, as shown in <figref idref="DRAWINGS">FIG. 11</figref>. After both feet are scanned, scanner <b>10</b> can determine the shoe size for the scanned feet.

0075Alternatively, after the feet are scanned/measured, the information is stored or used to make customized shoe lasts.

0076In accordance with another aspect of the present invention, each scanner <b>10</b> is provided with a communication device so that it can communicate with its home base. The communication device can be a telephone modem so that scanner(s) <b>10</b> communicates with home base through telephone lines. Alternatively, the communication device can be wireless, for example, similar to blackberry e-mail devices. After a scanner <b>10</b> establishes communication with home base, it can transfer information stored thereon relating to the scanned feet to home base, and it can receive commands or information from home base, for example, marketing information relating to the locality where scanner <b>10</b> is being deployed. The scanned feet information transferred to home base can be used to develop new shoe sizes to meet market needs.

0077In accordance with another aspect of the present invention, each scanner <b>10</b> has a unique IP address similar to the IP address of a desktop computer or laptop, so that home base can distinguish one scanner from another and that home base can send targeted information to different scanners.

0078While various descriptions of the present invention are described above, it is understood that the various features of the embodiments of the present invention shown herein can be used singly or in combination thereof. This invention is also not to be limited to the specifically preferred embodiments depicted therein.